rac N-Desmethyl Tramadol

• CAS: 147762-57-0
• Synonyms: (1R,2R)-(+)-1-(3-Methoxyphenyl)-2-[(methylamino)methyl]-cyclohexanol; _x000B_cis-(+)-1-(3-Methoxyphenyl)-2-[(methylamino)methyl]-cyclohexanol
• Molecular Formula: C₁₅H₂₃NO₂
• Molecular Weight: 249.35
• SMILES: CNCC1CCCCC1(C2=CC(=CC=C2)OC)O
• InChI: InChI=1S/C15H23NO2/c1-16-11-13-6-3-4-9-15(13,17)12-7-5-8-14(10-12)18-2/h5,7-8,10,13,16-17H,3-4,6,9,11H2,1-2H3/t13-,15+/m1/s1

rac N-Desmethyl Tramadol (NDT, also designated Metabolite M2) is the primary N-demethylated metabolite of the synthetic analgesic tramadol, formed via CYP3A4 and CYP2B6-mediated oxidation. Unlike the active O-desmethyltramadol (M1) metabolite, racemic NDT exhibits negligible affinity (Ki > 10 µM) for the cloned human µ-opioid receptor and demonstrates no functional agonist activity in [³⁵S]GTPγS binding assays.

rac N-Desmethyl Tramadol (CAS 147762-57-0): Essential Reference Standard for Tramadol Metabolic Pathway Dissection

rac N-Desmethyl Tramadol (NDT, also designated Metabolite M2) is the primary N-demethylated metabolite of the synthetic analgesic tramadol, formed via CYP3A4 and CYP2B6-mediated oxidation [1]. Unlike the active O-desmethyltramadol (M1) metabolite, racemic NDT exhibits negligible affinity (Ki > 10 µM) for the cloned human µ-opioid receptor and demonstrates no functional agonist activity in [³⁵S]GTPγS binding assays [2]. This compound exists as a racemic mixture of (+)- and (−)-enantiomers and serves as a critical analytical reference standard for forensic toxicology, clinical pharmacokinetic monitoring, and pharmacogenetic studies, where its distinct metabolic origin and enantioselective disposition enable applications that the pharmacologically active M1 metabolite cannot fulfill [3].

• [1] T'jollyn H, Snoeys J, Vermeulen A, et al. Physiologically Based Pharmacokinetic Predictions of Tramadol Exposure Throughout Pediatric Life: an Analysis of the Different Clearance Contributors with Emphasis on CYP2D6 Maturation. AAPS J. 2015;17(6):1376-1387. doi:10.1208/s12248-015-9803-z View Source
• [2] Gillen C, Haurand M, Kobelt DJ, Wnendt S. Affinity, potency and efficacy of tramadol and its metabolites at the cloned human µ-opioid receptor. Naunyn Schmiedebergs Arch Pharmacol. 2000;362(2):116-121. doi:10.1007/s002100000266 View Source
• [3] Haage P, Kronstrand R, Josefsson M, et al. Enantioselective pharmacokinetics of tramadol and its three main metabolites; impact of CYP2D6, CYP2B6, and CYP3A4 genotype. Pharmacol Res Perspect. 2018;6(4):e00419. doi:10.1002/prp2.419 View Source

Why rac N-Desmethyl Tramadol Cannot Be Replaced by O-Desmethyltramadol or Tramadol in Analytical and Pharmacogenetic Workflows

Substituting rac N-Desmethyl Tramadol with O-desmethyltramadol (M1), tramadol, or other in-class metabolites introduces fundamental errors in both quantitative and qualitative analytical determinations. The three compounds differ by orders of magnitude in µ-opioid receptor pharmacology: M1 binds with high affinity (Ki = 3.4 nM) and acts as a potent agonist, whereas NDT is essentially inert at this receptor (Ki > 10 µM, no GTPγS stimulation) [1]. Critically, their formation is mediated by distinct cytochrome P450 enzymes—M1 via CYP2D6, NDT via CYP3A4 and CYP2B6—meaning that genetic polymorphisms, drug-drug interactions, or disease states affecting one pathway may spare the other [2]. Furthermore, the NDT enantiomer ratio exhibits a uniquely large temporal shift (from ~2 to ~7 over 24 h) that is not replicated to the same degree by other tramadol metabolites, making NDT the preferred biomarker for time-since-ingestion estimation in forensic casework [3]. These orthogonal differences in pharmacology, metabolic origin, and enantioselective pharmacokinetics mean that no single comparator can simultaneously substitute for NDT across all its key application domains.

• [1] Gillen C, Haurand M, Kobelt DJ, Wnendt S. Affinity, potency and efficacy of tramadol and its metabolites at the cloned human µ-opioid receptor. Naunyn Schmiedebergs Arch Pharmacol. 2000;362(2):116-121. doi:10.1007/s002100000266 View Source
• [2] T'jollyn H, Snoeys J, Vermeulen A, et al. Physiologically Based Pharmacokinetic Predictions of Tramadol Exposure Throughout Pediatric Life: an Analysis of the Different Clearance Contributors with Emphasis on CYP2D6 Maturation. AAPS J. 2015;17(6):1376-1387. doi:10.1208/s12248-015-9803-z View Source
• [3] Haage P, Kronstrand R, Josefsson M, et al. Enantioselective pharmacokinetics of tramadol and its three main metabolites; impact of CYP2D6, CYP2B6, and CYP3A4 genotype. Pharmacol Res Perspect. 2018;6(4):e00419. doi:10.1002/prp2.419 View Source

Quantitative Differentiation Evidence: rac N-Desmethyl Tramadol vs. Closest Analogs

Mu-Opioid Receptor Binding Affinity: NDT Is Functionally Inert vs. High-Affinity M1

At the cloned human µ-opioid receptor expressed in CHO cell membranes, racemic N-desmethyltramadol (M2) exhibits a binding affinity (Ki) greater than 10 µM, indicating negligible receptor interaction [1]. In contrast, the primary active metabolite (+)-O-desmethyltramadol (M1) shows a Ki of 3.4 nM under identical assay conditions—a difference of approximately 3,000-fold [1]. For further context, (±)-tramadol itself exhibits an intermediate Ki of 2.4 µM, while (±)-M5 (N,O-didesmethyltramadol) binds with a Ki of 100 nM [1]. These data were generated by competitive inhibition of [³H]naloxone binding in membranes from stably transfected CHO cells [1].

Evidence Dimension	Binding affinity (Ki) at cloned human µ-opioid receptor
Target Compound Data	rac N-Desmethyl Tramadol (M2): Ki > 10,000 nM
Comparator Or Baseline	(+)-O-Desmethyltramadol (M1): Ki = 3.4 nM; (±)-Tramadol: Ki = 2,400 nM; (±)-M5: Ki = 100 nM
Quantified Difference	NDT Ki at least ~3,000-fold higher than (+)-M1; NDT Ki at least ~4-fold higher than tramadol
Conditions	Cloned human µ-opioid receptor, CHO cell membranes, [³H]naloxone competition binding

Why This Matters

The near-absence of µ-opioid binding establishes NDT as the appropriate negative control for opioid receptor pharmacology experiments, whereas M1 or tramadol cannot serve this purpose due to their measurable receptor engagement.

• [1] Gillen C, Haurand M, Kobelt DJ, Wnendt S. Affinity, potency and efficacy of tramadol and its metabolites at the cloned human µ-opioid receptor. Naunyn Schmiedebergs Arch Pharmacol. 2000;362(2):116-121. doi:10.1007/s002100000266 View Source

Functional Agonist Activity: NDT Fails to Stimulate GTPγS Binding vs. Potent Agonism by M1

In the [³⁵S]GTPγS functional assay using the same cloned human µ-opioid receptor system, racemic NDT and its congeners M3 and M4 produced no stimulatory effect on GTPγS binding, confirming a complete lack of agonist activity [1]. By contrast, (+)-M1, (−)-M1, and (±)-M5 all demonstrated measurable agonist activity with the following rank order of intrinsic efficacy: (+)-M1 > (±)-M5 > (−)-M1 [1]. The assay enables measurement of potency (EC₅₀), efficacy (Eₘₐₓ), and relative intrinsic efficacy as a function of receptor occupancy, providing a comprehensive functional profile [1].

Evidence Dimension	Functional agonist activity at cloned human µ-opioid receptor
Target Compound Data	rac N-Desmethyl Tramadol (M2): no stimulatory effect on [³⁵S]GTPγS binding
Comparator Or Baseline	(+)-M1: active agonist, highest intrinsic efficacy; (±)-M5: active agonist; (−)-M1: active agonist
Quantified Difference	NDT: zero detectable agonist activity vs. (+)-M1: robust agonist response
Conditions	[³⁵S]GTPγS binding assay, cloned human µ-opioid receptor, CHO cell membranes

Why This Matters

The absence of functional activity confirms NDT as the only major tramadol metabolite that can serve as a true inert comparator in opioid signaling studies; procurement of M1, M5, or tramadol would introduce confounding agonist effects.

• [1] Gillen C, Haurand M, Kobelt DJ, Wnendt S. Affinity, potency and efficacy of tramadol and its metabolites at the cloned human µ-opioid receptor. Naunyn Schmiedebergs Arch Pharmacol. 2000;362(2):116-121. doi:10.1007/s002100000266 View Source

Enantiomer Ratio Temporal Shift: NDT Provides a Unique Forensic Time-Since-Ingestion Biomarker

Among all four enantiomer pairs measured (tramadol, ODT, NDT, and NODT), the N-desmethyltramadol enantiomer ratio exhibited the largest temporal increase during the 24 hours following a single oral dose of tramadol (50 or 100 mg) in healthy volunteers [1]. In CYP2D6 extensive and intermediate metabolizers, the (+)/(−)-NDT enantiomer ratio rose from approximately 2 to nearly 7 over this period—a 3.5-fold increase [1]. While all four enantiomer pairs showed positive correlations with time post-administration, the magnitude of the NDT shift was substantially larger than that observed for tramadol, O-desmethyltramadol, or N,O-didesmethyltramadol enantiomer pairs [1]. This pronounced temporal shift was characterized using enantioselective LC-MS/MS quantification in whole blood samples collected at 17 timepoints over 24 hours [1].

Evidence Dimension	Change in (+)/(−)-enantiomer ratio over 24 h post-dose
Target Compound Data	NDT enantiomer ratio: ~2 → ~7 (3.5-fold increase) in CYP2D6 EM/IM subjects
Comparator Or Baseline	Tramadol, ODT, and NODT enantiomer ratios: all showed smaller temporal shifts (exact values not reported; NDT shift was the largest among all four pairs)
Quantified Difference	NDT ratio shift (from ~2 to ~7) is the largest among all measured analyte pairs
Conditions	Single oral dose (50 or 100 mg racemic tramadol), 19 healthy volunteers, 17 blood sampling timepoints over 24 h, enantioselective LC-MS/MS

Why This Matters

The uniquely large NDT enantiomer ratio shift makes it the preferred biomarker for estimating time since tramadol ingestion in forensic and clinical toxicology casework—an application where ODT, tramadol, or NODT provide inferior temporal resolution.

• [1] Haage P, Kronstrand R, Josefsson M, et al. Enantioselective pharmacokinetics of tramadol and its three main metabolites; impact of CYP2D6, CYP2B6, and CYP3A4 genotype. Pharmacol Res Perspect. 2018;6(4):e00419. doi:10.1002/prp2.419 View Source

Divergent CYP Enzyme Dependency: NDT Formation via CYP3A4/CYP2B6 Is Orthogonal to M1 Formation via CYP2D6

In vitro studies using pediatric human liver microsomal (HLM) batches demonstrated that O-desmethyltramadol (M1) formation was mediated exclusively by CYP2D6, whereas N-desmethyltramadol (NDT) formation was mediated—at least in part—by CYP3A4 [1]. These findings were incorporated into a physiologically based pharmacokinetic (PBPK) model in Simcyp®, where adult clearance values were qualified for CYP2D6, CYP3A4, CYP2B6, and renal contributions [1]. The distinct enzymatic routes mean that CYP2D6 genetic polymorphisms (poor, intermediate, extensive, or ultrarapid metabolizer status) predominantly affect M1 formation, while NDT formation is more sensitive to CYP3A4 and CYP2B6 activity [2]. This metabolic independence is further evidenced by the accumulation of NDT enantiomers in CYP2D6 poor metabolizers, who exhibit 7.40- and 8.69-fold greater mean AUC values for M2 enantiomers compared to extensive metabolizers, due to impaired secondary CYP2D6-mediated metabolism of NDT [3].

Evidence Dimension	Primary cytochrome P450 enzyme responsible for metabolite formation
Target Compound Data	NDT: formation mediated by CYP3A4 and CYP2B6 (in vitro HLM data); AUC 7.40–8.69-fold higher in CYP2D6 PMs vs. EMs
Comparator Or Baseline	ODT (M1): formation mediated exclusively by CYP2D6; AUC 4.33-fold higher in CYP2D6 EMs vs. PMs
Quantified Difference	NDT formation: CYP3A4/CYP2B6-dependent; M1 formation: CYP2D6-exclusive; NDT AUC PM/EM ratio = 7.40–8.69 vs. M1 AUC EM/PM ratio = 4.33
Conditions	In vitro: pediatric HLM batches (1 and 3 months); In vivo: single oral 100 mg tramadol, Spanish healthy volunteers, CYP2D6-phenotyped (5 PMs, 19 EMs)

Why This Matters

The orthogonal CYP dependency allows NDT to serve as a functional probe for CYP3A4/CYP2B6 activity independent of CYP2D6 status, enabling multi-pathway phenotyping from a single tramadol administration—a capability unavailable when measuring only M1 or the parent drug.

• [1] T'jollyn H, Snoeys J, Vermeulen A, et al. Physiologically Based Pharmacokinetic Predictions of Tramadol Exposure Throughout Pediatric Life: an Analysis of the Different Clearance Contributors with Emphasis on CYP2D6 Maturation. AAPS J. 2015;17(6):1376-1387. doi:10.1208/s12248-015-9803-z View Source
• [2] Haage P, Kronstrand R, Josefsson M, et al. Enantioselective pharmacokinetics of tramadol and its three main metabolites; impact of CYP2D6, CYP2B6, and CYP3A4 genotype. Pharmacol Res Perspect. 2018;6(4):e00419. doi:10.1002/prp2.419 View Source
• [3] García-Quetglas E, Azanza JR, Sádaba B, Muñoz MJ, Gil I, Campanero MA. Pharmacokinetics of tramadol enantiomers and their respective phase I metabolites in relation to CYP2D6 phenotype. Pharmacol Res. 2007;55(2):122-130. doi:10.1016/j.phrs.2006.11.006 View Source

Plasma Pharmacokinetics: NDT Achieves Lower and Later Cmax vs. M1 Following Oral Tramadol

In a pharmacokinetic study of 24 healthy volunteers receiving a single 100 mg oral dose of racemic tramadol, the mean maximum plasma concentration (Cmax) of N-desmethyltramadol (M2) was 23.1 ± 11.4 ng/mL, achieved at a Tmax of 2.8 ± 1.0 hours [1]. For comparison, O-desmethyltramadol (M1) achieved a Cmax of 88.7 ± 30.3 ng/mL at a Tmax of 2.4 ± 0.7 hours—a 3.8-fold higher peak concentration occurring approximately 0.4 hours earlier [1]. Tramadol itself reached a Cmax of 349.3 ± 76.7 ng/mL at 1.6 ± 0.4 hours [1]. The renal clearance values further differentiated the metabolites: M1 = 193.9 ± 67.6 mL/min, M2 = 116.1 ± 61.8 mL/min, and M5 = 252.0 ± 91.5 mL/min [1]. Urinary recovery over 30 hours accounted for 11.2% of the dose as M1 but only 1.1% as M2 [1].

Evidence Dimension	Maximum plasma concentration (Cmax) after single 100 mg oral tramadol
Target Compound Data	NDT (M2): Cmax = 23.1 ± 11.4 ng/mL; Tmax = 2.8 ± 1.0 h; renal clearance = 116.1 ± 61.8 mL/min; urinary recovery = 1.1%
Comparator Or Baseline	ODT (M1): Cmax = 88.7 ± 30.3 ng/mL; Tmax = 2.4 ± 0.7 h; renal clearance = 193.9 ± 67.6 mL/min; urinary recovery = 11.2%; Tramadol: Cmax = 349.3 ± 76.7 ng/mL
Quantified Difference	NDT Cmax ~3.8-fold lower than ODT; NDT Tmax ~0.4 h later; NDT renal clearance ~1.7-fold lower; NDT urinary recovery ~10-fold lower
Conditions	24 healthy volunteers (12 male, 12 female), single oral 100 mg racemic tramadol, plasma analyzed by HPLC, 30 h urine collection

Why This Matters

The distinct PK profile of NDT—lower Cmax, later Tmax, and markedly lower urinary recovery—requires that any analytical method validated for tramadol or M1 be separately validated for NDT quantification; procurement of a dedicated NDT reference standard is essential for accurate bioanalytical method development.

• [1] Ardakani YH, Rouini MR. Pharmacokinetic study of tramadol and its three metabolites in plasma, saliva and urine. DARU J Pharm Sci. 2009;17(4):245-255. View Source

Analytical Method Performance: NDT LOQ Is 2-Fold Higher Than M1 in Validated GC-MS Urine Assays

In a validated GC-MS method for the simultaneous determination of tramadol (T), O-desmethyltramadol (ODT), and N-desmethyltramadol (NDT) in human urine, the respective limits of quantification (LOQ) were 10, 10, and 20 ng/mL [1]. The extraction efficiency for NDT (98.21%) was slightly lower than that for ODT (101.30%) and tramadol (102.12%) using liquid-liquid extraction with methyl-tert-butyl ether followed by back-extraction with 0.1 M HCl [1]. Intra-assay precision ranged from 1.29% to 6.48% and inter-assay precision from 1.28% to 6.84% across all three analytes [1]. Intra-assay accuracy was within 91.79–106.89% for all analytes [1]. The method successfully detected urine concentrations of all three analytes in six healthy volunteers over 7 days following a 50 mg oral tramadol dose, but the inherently lower abundance and higher LOQ of NDT necessitate careful method optimization distinct from that required for tramadol or M1 [1].

Evidence Dimension	Limit of quantification (LOQ) in human urine GC-MS method
Target Compound Data	NDT: LOQ = 20 ng/mL; extraction efficiency = 98.21%
Comparator Or Baseline	Tramadol: LOQ = 10 ng/mL, extraction efficiency = 102.12%; ODT: LOQ = 10 ng/mL, extraction efficiency = 101.30%
Quantified Difference	NDT LOQ is 2-fold higher (worse) than tramadol and ODT LOQ values
Conditions	GC-MS method, human urine, liquid-liquid extraction with MTBE, back-extraction with 0.1 M HCl, Proadifen (SKF525A) internal standard, linear range 10–1000 ng/mL (r² > 0.99)

Why This Matters

The 2-fold higher LOQ for NDT relative to tramadol and ODT means that analytical methods validated only for the parent drug or M1 may fail to detect low-concentration NDT in forensic or clinical samples; dedicated NDT reference standards are required for proper method validation and routine use.

• [1] El-Sayed AY, Mohamed KM, Nasser AY, Button J, Holt DW. Simultaneous determination of tramadol, O-desmethyltramadol and N-desmethyltramadol in human urine by gas chromatography-mass spectrometry. J Chromatogr B. 2013;926:9-15. doi:10.1016/j.jchromb.2013.02.019 View Source

Procurement-Driven Application Scenarios for rac N-Desmethyl Tramadol (CAS 147762-57-0)

Forensic Toxicology: Time-Since-Ingestion Estimation via Enantiomer Ratio Analysis

Forensic toxicology laboratories requiring estimation of time elapsed since tramadol ingestion should procure rac N-Desmethyl Tramadol as the primary biomarker reference standard. The compound's unique (+)/(−)-enantiomer ratio shift—from approximately 2 to nearly 7 over 24 hours in CYP2D6 extensive/intermediate metabolizers—provides the largest temporal dynamic range among all tramadol metabolites [1]. This evidence, established through enantioselective LC-MS/MS analysis of whole blood samples collected at 17 serial timepoints following controlled oral tramadol administration, supports the use of NDT enantiomer ratios to distinguish recent from past tramadol intake in postmortem and criminal casework [1]. Procurement of enantiopure or racemic NDT reference material is essential for calibrating the chiral chromatographic methods required to implement this forensic technique.

Clinical Pharmacogenetic Phenotyping: CYP2D6 Activity Assessment via NDT Accumulation

Clinical pharmacology laboratories conducting CYP2D6 phenotyping studies should include N-desmethyltramadol as a key analyte in their analytical panel. In CYP2D6 poor metabolizers (PMs), NDT enantiomers accumulate to AUC values 7.40- to 8.69-fold greater than those observed in extensive metabolizers, providing a sensitive and quantitative readout of impaired CYP2D6-mediated secondary metabolism [2]. This metabolic signature is orthogonal to the O-desmethyltramadol (M1) formation pathway (CYP2D6-dependent) and the NDT formation pathway (CYP3A4/CYP2B6-dependent), enabling multi-enzyme activity assessment from a single tramadol probe administration [3]. Laboratories performing pharmacogenetic-guided tramadol dosing or drug-drug interaction studies require certified NDT reference standards for accurate enantioselective quantification in plasma or whole blood.

Opioid Receptor Pharmacology: Definitive Negative Control for µ-Opioid Functional Assays

Academic and pharmaceutical researchers studying µ-opioid receptor pharmacology should procure racemic N-desmethyltramadol as a structurally matched negative control compound. NDT exhibits a binding affinity (Ki) greater than 10 µM at the cloned human µ-opioid receptor—at least 3,000-fold weaker than (+)-O-desmethyltramadol (Ki = 3.4 nM)—and demonstrates zero stimulatory effect in the [³⁵S]GTPγS functional assay, confirming complete absence of agonist activity [4]. Other tramadol metabolites (M1, M5) and the parent drug all display measurable µ-opioid receptor engagement and cannot serve this purpose. For structure-activity relationship (SAR) studies investigating the molecular determinants of tramadol-derived opioid activity, NDT provides a critical baseline that controls for non-opioid chemical effects while eliminating confounding opioid receptor activation.

Bioanalytical Method Development: Multi-Analyte LC-MS/MS and GC-MS Validation

Bioanalytical laboratories developing validated methods for the simultaneous quantification of tramadol and its metabolites in biological matrices must include N-desmethyltramadol in their analyte panel. The validated GC-MS method for human urine demonstrates that NDT requires a distinct limit of quantification (LOQ = 20 ng/mL, versus 10 ng/mL for tramadol and ODT) and exhibits a slightly lower extraction efficiency (98.21%) compared to the other analytes [5]. These performance differences, combined with NDT's lower plasma concentrations (Cmax = 23.1 ng/mL, versus 88.7 ng/mL for M1), mean that method sensitivity and linearity must be specifically established for NDT rather than extrapolated from tramadol or M1 validation data [6]. Procurement of a certified NDT reference standard of known purity is mandatory for preparing calibration standards and quality control samples in compliance with regulatory bioanalytical method validation guidelines.

• [1] Haage P, Kronstrand R, Josefsson M, et al. Enantioselective pharmacokinetics of tramadol and its three main metabolites; impact of CYP2D6, CYP2B6, and CYP3A4 genotype. Pharmacol Res Perspect. 2018;6(4):e00419. doi:10.1002/prp2.419 View Source
• [2] García-Quetglas E, Azanza JR, Sádaba B, Muñoz MJ, Gil I, Campanero MA. Pharmacokinetics of tramadol enantiomers and their respective phase I metabolites in relation to CYP2D6 phenotype. Pharmacol Res. 2007;55(2):122-130. doi:10.1016/j.phrs.2006.11.006 View Source
• [3] T'jollyn H, Snoeys J, Vermeulen A, et al. Physiologically Based Pharmacokinetic Predictions of Tramadol Exposure Throughout Pediatric Life: an Analysis of the Different Clearance Contributors with Emphasis on CYP2D6 Maturation. AAPS J. 2015;17(6):1376-1387. doi:10.1208/s12248-015-9803-z View Source
• [4] Gillen C, Haurand M, Kobelt DJ, Wnendt S. Affinity, potency and efficacy of tramadol and its metabolites at the cloned human µ-opioid receptor. Naunyn Schmiedebergs Arch Pharmacol. 2000;362(2):116-121. doi:10.1007/s002100000266 View Source
• [5] El-Sayed AY, Mohamed KM, Nasser AY, Button J, Holt DW. Simultaneous determination of tramadol, O-desmethyltramadol and N-desmethyltramadol in human urine by gas chromatography-mass spectrometry. J Chromatogr B. 2013;926:9-15. doi:10.1016/j.jchromb.2013.02.019 View Source
• [6] Ardakani YH, Rouini MR. Pharmacokinetic study of tramadol and its three metabolites in plasma, saliva and urine. DARU J Pharm Sci. 2009;17(4):245-255. View Source